The overall goal of this project is to examine whether cognitive status and cognitive change in middle age are predictive of rate of subsequent decline. While stability is the normative pattern for most mental abilities in midlife, longitudinal studies of normal aging have reported subgroups of middle-aged adults who show reliable decline or improvement on specific cognitive abilities. Adults who exhibit early decline in midlife on select abilities may be at higher risk for an accelerated rate of cognitive decline in old age. Study participants (N = 653) will be members from the Seattle Longitudinal Study (SLS) with longitudinal cognitive data obtained during midlife. Midlife cognitive change data will be used to identify patterns of change and to develop Midlife Cognitive Risk criteria for three domains: Episodic memory, Executive functioning, and Processing speed. The impact of midlife cognition will be studied for two cohorts. For the Older Cohort (b1914 -1941; M age = 75; N = 332) cognitive data are available in midlife and also in old age. For the Middle Age Cohort (b1942 -1962; N = 321; M age = 55) cognitive change data are available in midlife. The sample will be further characterized by neuropsychological assessment, APOE genotyping, and health histories. To examine the association of brain volume and cognitive change, baseline measures and rate of atrophy in the hippocampus, entorhinal cortex.prefrontal cortex and frontal white matter will be assessed by structural MRI at two occasions. Longitudinal data on participants' engagement in stimulating cognitive activities will also be studied as a mediator of cognitive change.